616
Biochemistry
P. R. SRINIVASAN AND ERNEST BOREK
communication, 1963). Thus we begin to have a better understanding of the Cotton effects of nucleic acids. In a preliminary communication (Yang and Samejima, 1963b) we observed that dCMP will dominate the 290 mp peak in a mixture of deoxyribonucleotides. Therefore, a close relationship may exist between the (cytosine guanine) content and the 290 mp peak for DNA when its secondary structure is destroyed; that is, the higher the (C G) content, the larger the magnitude of the peak. Of course, the ORD of the mononucleotides could vary not only in magnitude but even in sign when they are incorporated into a polynucleotide chain. This hypothesis could be tested by studying the Cotton effects of DNA of various species. Furthermore, a comparative study of the Cotton effects of the DNA’s in their native state might provide some information concerning the stacking interactions of the base pairs in the different species. ORD of RNA.-Figure 4 suggests that RNA also possesses a certain degree of secondary structure and that its Cotton effects are significantly changed in alkaline solution or a t high temperature. The red shift of the 280 mp peak and 252 mp trough (curves 2 and 3) might indicate the disruption of intramolecular hydrogen bonding between the stacking chromophores (bases). On the other hand, the weak 222 mp peak actually increases its magnitude in alkaline solution or a t high temperature. Why the peaks shift in opposite directions is not clear. Of interest is the similarity of the ORD curve for DNA (Fig. 2, curve 4) and RNA (Fig. 4, curve 3) in alkaline solution. DNA and RNA also appear to have very similar strong peaks a t 200 mp and 195 mp, respectively, with nearly identical magnitude (Figs. 2 and 4). Since we are still unable to measure this Cotton effect a t high temperature because of experimental difficulties, it is
+
+
not certain whether this peak is closely related to the secondary structure or not. Note also that the measurements below 200 mp approach the instrument limit, and therefore should be viewed with some reservation until we have more concrete evidence by studying the Cotton effects of various nucleic acids. ACKNOWLEDGMENT We thank Professor J. R. Fresco for a frozen solution of calf thymus DNA, Professor E. P. Geiduschek for his generous gift of salmon sperm DNA, and Dr. T. T. Herskovits for the rat liver RNA solution. REFERENCES Brahms, J. (1963), J. A m . Chem. SOC.85, 3298. DeVoe, H., and Tinoco, I., Jr. (1963), J. Mol. Biol. 4, 518. Fresco, J. R. (1961), Tetrahedron 13, 185. Fresco, J. R., Lesk, A. M., Gorn, R., and Doty, P. (1961), J . A m . Chem. SOC.83,3155. Harris, T. L., Hirst, E. L., and Wood, C. E. (1932), J. Chem. SOC.,2108. Kasha. M. (1961). in Liaht and Life. McElrov. W. D.. and Glass, B.,‘ eds.; ‘Baltimore, Johns Hopkins Press, p.’ 31. Rich, A., and Kasha, M. (1960), J . A m . Chem. SOC.82, 6197. Simmons, N. S., and Blout, E. R. (1960), Biuphys. J. 1, 55. Tinoco, I., Jr. (1964), J. A m . Chem. SOC.86,297. Ts’o, P. 0. P., Helmkamp, G. K., and Sander, C. (1962), Biochim. Biophys. Acta 55, 584. Urnes, P., and Doty, P. (1961), Aduan. Protein Chem. 16, 401. Voet, D., Gratzer, W. B., Cos, R. A., and Doty, P. (1963), Biopolymers 1, 193. Yang, J. T., and Samejima, T. (1963a), J . Biol. Chem. 238, 3262, Yang, J. T., and Samejima, T. (1963b), J. A m . Chem. SOC. 85,4039.
Species Variation of the RNA Methylases* P. R. SRINIVASAN AND ERNEST BOREK From the Department of Biological Chemistry, College of Physicians and Surgeons, Columbia University, New York Received January 2, 1964 The complex of enzymes which methylate t-RNA a t the polynucleotide level were found to be species specific. Therefore a preparation of t-RNA (transfer ribonucleic acid) from a given source, while fully methylated with respect to its homologous enzymes, will receive methyl groups in vitro from a heterologous enzyme source. Some closely related organisms exhibit a similarity in their pattern of methylation of foreign t-RNA’s. A study of amino acid-specific t-RNA’s with heterologous enzymes revealed that the different t-RNA’s do not behave uniformly as substrates for methylation. Therefore the species variation may be restricted only to some of the t-RNA’s.
It has recently been demonstrated (Borek et al., 1962; Fleissner and Borek, 1963) and confirmed in several laboratories (Svensson et al., 1963; Starr, 1963; Gold et al., 1963; Comb, 1963) that the methylated bases of t-RNA are acquired by the methylation of the preformed polynucleotide chain by a complex of enzymes. The demonstration of the sequential nature of t-RNA synthesis was made possible by the * T h i s work was supported by grants (E-4671, 10384, and 1181) from the National Institutes of Health, U. S. Public Health Service, and a grant (G-22140) from the National Science Foundation, as well as by a contract (AT(30-1)2358) from the U. S. Atomic Energy Commission.
discovery of a mutant, Escherichia coli KI2W6,in which, as a result of two independent mutations, the primary synthesis and the methylation of the product are uncoupled (Borek et al., 1955). During methionine starvation this organism produces a species of t-RNA devoid of methylated derivatives of the component bases (Mandel and Borek, 1963). The availability of a naturally produced nonmethylated t-RNA as a substrabe made possible the demonstration of the existence of a t-RNA-methylating enzyme. The enzyme system proved to be a complex whose individual members are endowed with highly restricted specificities (Fleissner and Borek, 1963; Gold et al., 1963).
Vol. 3, No. 5, M a y , 1964 The nonmethylated t-RNA, when isolated from the prestarved methionine auxotroph, can accept methyl groups not only from homologous enzymes but from enzymes extracted from every other species of organism tested. In studies of the interactions of heterologous enzymes and methyl-deficient t-RNA control experiments were performed using fully methylated t-RNA from E . coli K12 in logarithmic growth phase. Instead of the anticipated negligible interactions, extensive transfer of supernumerary methyl groups into the otherwise fully methylated t-RNA by several heterologous enzymes was observed. Such overmethylation of normal t-RNA by heterologous enzymes is not restricted to RNA from E . coli, but is a rather generalized phenomenon: successful interactions of heterologous RNA and enzyme are far more frequent than negative findings. The distribution of methyl groups in t-RNA is apparently a species characteristic attribute; therefore an RNA, while fully methylated with respect to its own enzyme, can offer new sites for methylation to heterologous enzymes (Srinivasan and Borek, 1963). The species specificity of RNA methylases has been observed in other laboratories as well (Svensson etal., 1963; Gold etal., 1963). It has been suggested by Kornberg and his associates (Kornberg et al., 1959) that the 5-methylcytosine in germinating wheat DNA may be acquired by enzymatic methylation of the preformed polymer. The absence of a nonmethylated substrate DNA for a homologous enzyme is a handicap for the search for a DNA methylase. However, the discovery of species variation of RNA methylase made the species variation of DNA methylases even more likely since the distribution of methylated bases in DNA is even kingdom specific: animals and plants contain 5methylcytosine, bacteria contain 6-methyladenine. The species specificity of DNA methylases has been demonstrated in Dr. Hurwitz's laboratory and in our own (Gold et al., 1963; P. R. Srinivasan and E. Borek, in preparation). We report here detailed studies on the species specificity of the RNA methylases. MATERIALS AND METHODS The bacterial strains used in this investigation were all ultimately derived from stocks of the American Type Culture Collection. The organisms were grown on nutrient broth except for E . coli KI2, which was grown on defined media. All the organisms were harvested in logarithmic growth phase. For the preparation of methyl-deficient t-RNA see Mandel and Borek (1963). The enzyme extract from yeast wzs prepared from commercially available baker's yeast. E . coli B and yeast t-RNA's were purchased from General Biochemicals, Inc., Chagrin Falls, Ohio. S-Adenosyl-~-methionine-methyl-~4C with a specific activity of 26.11 mc/mM was obtained as an aqueous acid solution from Tracerlab, Inc., Waltham, Mass. It w?s neutralized with 1 M Tris to p H 5.0 and diluted to yield an Rctivitg of 10 pc/ml. Preparation of Enzyme Extracts and t-RNA.The prepsration of liver and spinach extracts hss been described in detgil in an earlier public?.t-ion(Srinivasan and Borek, 1963). To obtain the bacterial and yeastenzyme extracts the cells were disrupted in a Waring Blendor with glass beads; the cell debris and unbroken cells were eliminated by centrifugation a t 20,000 g for 20 minutes, and a clear extract was obtained by centrifugation a t 105,000 g in the Spinco preparative ultracentrifuge for 1 hour. This procedure is similar to that described by Chamberlin and Berg (1962)
617
SPECIES VARIATION OF THE RNA METHYLASES
and is described in detail by Fleissner and Borek (1963). The extracts were stored at 2 " with a few drops of toluene and under these conditions they remained active for a t least a week. The t-RNA preparations from each strain were prepared by phenol extractions of the supernatant which results from centrifugation a t 105,000 g for 2 hours. The RNA was precipitated by alcohol, dialyzed, and lyophilized. The reaction mixtures contained 100 pmoles of Tris buffer, p H 8.2, 10 pmoles each of MgC12 and reduced glutathione, 0.05 ml of S-adenosylmethioninemethyl-lC, 2 mg of t-RNA, and 1 ml of enzyme extract in a total volume of 2.0 ml. The incubations were carried out a t 37 for 45 minutes. The precipitation of t-RNA from the reaction mixture and the subsequent washing procedure to remove contamination of adenosylmethionine have been described earlier (Srinivasan and Borek, 1963). With each extract a control experiment was performed to assess the incorporation in the absence of any added t-RNA. Such background values were subtracted from the results presented in the various tables. All the radioactivity measurements were carried out in a Nuclear Chicago gas-flow counter and sufficient counts were recorded to give a probable error of =t5yo. O
RESULTSAND DISCUSSION I n Table I the results of interactions of enzymes from liver, spinach, and yeast with t-RNA from a variety of sources are presented. In any attempt a t interpretation it must be borne in mind that these data are the results of a highly complex system of interactions. The substrates are the different amino acid-specific t-RNA's, which number over twenty, and several different RNA methylases have been identified to date (in E . coli). TABLE I
INTERACTION OF HETEROLOGOUS t-RNA's METHYLASES" Source of t-RNA Liver Spinach Yeast E. coli K,,
AND
RNA
Source of Enzyme Extract Liver
Spinach
Yeast
0.01 0.03 0.08 0.21
0 0.01 0.04 0.12
0.13 0.07 0.07 1.23
0.13 0.15 0.12 0.17 0.13 0.11 0.21
0.08 0.09 0.08 0.06 0.06 0.09 0.08
0.58 0.72 0.50 0.24 0.46 0.55 0.76
(methyl deficient) E. coli K,? log E. coli B B. megatherium B . cereus 'Ps.aerugenosa Rh. spheroides S. typhimurium
a Activity is expressed as mpmoles 14CH8 incorporated into 2 mg t-RNA with I4CH3S-adenosylmethionine as the methyl donor.
The following generalizations may be drawn. The methylation of the isolated t-RNA by the appropriate homologous enzyme system is, as a general rule, negligible except by the enzymes from yeast which consistently produce a small but significant incorporation. The reason for this is obscure a t the present time. It is possible that in the RNA extracted from yeast a small fraction is in a transient state between the primary synthesis and methylation. The existence of such t-RNA, transiently denuded of methyl groups,
618
Biochemistry
P. R. SRINIVASAN AND ERNEST BORER
TABLEI1 INTERACTION OF BACTERIAL t-RNA’s AND RNA METHYLASES
-
Source of t-RNAa Source of Enzyme Extract
K12
K12
B.
B. mega-
(log)
E. coli B
S. typhi-
(starved)
murium
cereus
therium
E. coli K12 E. coli B S. typhimurium B. cereus B. megatherium Ps. aerugenosa Ps. fluorescens Rh. spheroides
2.6 2.4 2.8 1.2 1.4 1.35 2.0 0.39
0.01 , 01 0 0.29 0.50 0.01 0.50
